Tech Specs

Reference.

Everything an engineer needs to embed, deploy, scale, and operate DocumentForge — organised by topic so you can jump straight to the section you need.

Core engine

Working with the database itself — querying, modelling JSON, getting started from source.

QL

Query Language →

The full SQL surface: SELECT, WHERE, JOIN, GROUP BY, ORDER BY, aggregations, dot-notation paths, array indexing, IN, DISTINCT, parameterised queries.

DM

Data Modeling →

When to embed, when to reference. Schema-on-read patterns, indexing strategy, composite keys, and how to keep nested arrays queryable.

GS

Library API →

Library-first guide for .NET developers: DocumentForgeDb, the LINQ surface, indexes, the REPL, and crash-recovery semantics.

Distribution & scale

Going beyond a single file — replication, sharding, multi-datacenter operations.

RP

Replication →

Logical replication with monotonic sequence numbers, follower catchup, auto-failover on leader silence, and zero-data-loss planned handover for datacenter moves.

SH

Sharding →

Consistent-hash routing, replicated reference collections (joins stay local), online rebalance with no maintenance window, and adding/dropping shards safely.

TX

Transactions →

Multi-document atomicity on a single node and across shards via two-phase commit. Crash-safe with automatic recovery; PREPARE timeout for liveness; same API on single-node and cluster handles.

DP

Deployment →

Running DocumentForge on bare metal, in containers, behind a load balancer, or via the included Render blueprint. Multi-shard topologies and config patterns.

Operations

Securing it, observing it, and driving it from the command line.

SE

Security →

API keys, replication secrets, TLS, threat model, what's protected and what isn't. Designed for airline-grade deployments where the airline owns the keys.

CLI

CLI Reference →

Every dfdb subcommand: serve, repl, query, seed, cluster, health, rebalance. Flags, examples, and exit codes.

PERF

Performance →

Full benchmark methodology, real numbers at 250K and 10M docs, the bottlenecks at each scale, and the tuning knobs that matter.

Looking for a higher-level view? Start with Use Cases for the why, or Quickstart if you'd rather skip ahead and run the binary.